No abstract

Identification of Public Bus Stops and characterising their waiting time is of significance in the domain of Intelligent Transportation System and public interest. Surveys suggest that the existing systems and algorithms cannot be merely replicated in the context of developing regions, because of the kind of heterogeneity and chaotic situation prevalent in their transport system, in contrast to the orderly and much regulated system of the developed world. Moreover a significant fraction of population residing here is not so tech-savvy, hence crowd-sourced based strategies renders useless. Either novel algorithms/techniques are required or the existing ones have to undergo suitable adaptation and innovation to suit the contrasting characteristics of developing regions. Using customized hardware we present the stoppage pattern analysis of public bus GPS traces of more than 2000 km for a 20 km route of Durgapur, a suburban city (West Bengal, India). We have revealed that 53.03 % of the travel time for the selected route has higher predictability with Standard deviation of 30.73% .

Machine learning models deployed as a service (MLaaS) are susceptible to model stealing attacks, where an adversary attempts to steal the model within a restricted access framework. While existing attacks demonstrate near-perfect clone-model performance using softmax predictions of the classification network, most of the APIs allow access to only the top-1 labels. In this work, we show that it is indeed possible to steal Machine Learning models by accessing only top-1 predictions (Hard Label setting) as well, without access to model gradients (Black-Box setting) or even the training dataset (Data-Free setting) within a low query budget. We propose a novel GAN-based framework 1 that trains the student and generator in tandem to steal the model effectively while overcoming the challenge of the hard label setting by utilizing gradients of the clone network as a proxy to the victim's gradients. We propose to overcome the large query costs associated with a typical Data-Free setting by utilizing publicly available (potentially unrelated) datasets as a weak image prior. We additionally show that even in the absence of such data, it is possible to achieve state-ofthe-art results within a low query budget using synthetically crafted samples. We are the first to demonstrate the scalability of Model Stealing in a restricted access setting on a 100 class dataset as well.